Business entities collect, store and use significant amounts of data on a day-to-day basis. Quick and ready access to that data drives transaction-related applications and operations of the business entity. An enterprise-class business entity can collect hundreds to thousands of new and updated business records from across the entity every minute. Those new and updated records can be entered into one or more data stores accessible to users throughout the enterprise. Users can access that data by using direct calls to the data store or through searches of the data store or an associated search index.
In order to optimize speed and performance in finding relevant records for a search query, records in a data store can be processed for a variety of information that can be entered into a search index. Without an index, a search process performing a search would have to scan every record in a data store, which can consume significant time and computing resources. A search engine index, on the other hand, collects, parses, and stores data in a manner to facilitate fast and accurate information retrieval.
In an enterprise-class business entity having a large number of new entries and update entries being made to a data store, there can be a significant delay between a time of entry of the new or updated record into the data store and a time when information related to that new or updated record is merged into the search index associated with the data store. During that period, a search of a search index for the new or updated information will yield no or outdated results. In such an instance, a lengthy and time consuming search of the data store for the new or updated record would have to be performed. Such a search of a large enterprise-wide data store can take many minutes to several hours, depending upon the size of the data store and the nature of the search. In an environment in which a search is being conducted in response to, for example, a telephonic customer query of a customer service representative, such a lengthy delay in response time can be unacceptable.
It is therefore desirable to have a system in which new or updated records to a data store can be quickly searched during a time period prior to information related to those records being entered into a search index.